Food insecurity and HIV/AIDS are two leading causes of morbidity and mortality in sub-Saharan Africa and are inextricably linked. Since food insecurity contributes to increased HIV transmission risk and higher HIV-related morbidity and mortality, the WHO, UNAIDS, and the World Food Programme have recommended integrating sustainable food production strategies into HIV/AIDS programming. Yet, to date there have been few studies to systematically evaluate the impact of promising food security interventions on health, economic and behavioral outcomes among people living with HIV and AIDS. To address this gap, we plan to test the hypothesis that a multi-sectoral agricultural intervention delivered in Nyanza Province, Kenya will prevent highly active antiretroviral (HAART) treatment failure, reduce co-morbidities, and decrease secondary HIV transmission risk. The intervention will include: a) a human-powered water pump and other required farm commodities, b) a microfinance loan (~$75) to purchase the pump and agricultural implements, and c) education in sustainable farming practices. To develop our intervention, we have formed an interdisciplinary collaboration with organizations in the healthcare, agriculture, and microfinance sectors. To improve our likelihood of success in a future cluster randomized control trial (RCT), we propose three specific aims for a 3-year R34 pilot study. In Aim 1 we will operationalize and pilot test key design elements of a future cluster RCT aimed to improve health outcomes among HAART-treated patients in Western Kenya. In conjunction with our collaborating partners, we will develop the different components of the intervention, including: a) randomization procedures for cluster RCT using detailed site assessments; b) agricultural training protocols; c) procedures for control group; d) manual of operations. In Aim 2 we will conduct a pilot study of an agricultural intervention to determine the preliminary impact of the intervention on mediating outcomes (food security, and household economic indicators), and on primary health outcomes of interest for the planned RCT (HIV treatment outcomes, HIV transmission risk and women's empowerment). One hundred and twenty HIV-infected farmers on HAART in Western Kenya (60 at an intervention clinic and 60 at a control clinic) will be enrolled and followed for 1 year. Impacts of our intervention on mediating and primary health outcomes will be investigated separately and jointly to provide a preliminary assessment of possible direct and indirect intervention effects. In Aim 3 we will assess the acceptability and feasibility of intervention and control conditions, and systematically translate lessons learned in the pilot study into the design of a cluster RCT. To accomplish Aim 3, we will conduct a mixed methods process evaluation of the different intervention components and their implementation using quantitative, qualitative, and observational methods. We will prepare an R01 grant for submission based on lessons learned. The ultimate goal of this work is to contribute to sustainable solutions to tackle the intersecting challenges of food insecurity, poverty, and HIV/AIDS in sub-Saharan Africa.